Several objective statistical techniques have been used to isolate zones of distinctive chemical characteristics within the saline facies of the Eocene Green River Formation. The data consist of measured concentrations of Al, Si, Na, K, Ca, Sr, Fe, and S in 374 composite samples collected at 0.61-m intervals in a 232-m section of the saline facies from a core of the Green River Formation, Piceance Creek basin, Colorado. The techniques were also applied to the oil-yield (Fischer assay) data for the entire core. The first statistical technique used was analysis-of-variance zonation. Zone boundaries determined by this technique separate portions of the core that have maximum between-zone variance and minimum within-zone variance. The second technique involves computation of oving correlation coefficients between two variables over successive 31-sample intervals. This technique isolates zones of strong positive and negative associations among geochemical variables.

On the basis of these two statistical techniques, the saline facies and underlying Garden Gulch Member of the Green River Formation have been divided into six geochemical zones. The most useful variables for zoning are Si, Al, Fe, S, Na, and oil yield. Concentrations of silicon and aluminum exhibit considerable variation and are positively correlated throughout the saline facies, expressing a strong correlation between quartz and dawsonite. The minerals most characteristic of the saline facies of the Green River Formation are dawsonite and nahcolite. Concentrations of iron and sulfur (present mainly as pyrite and marcasite) are positively correlated with oil yield in zones of higher salinity (as indicated by zones containing highest concentrations of nahcolite) and negatively correlat d with oil yield in zones of lower salinity. This suggests that the effect of organic content on iron diagenesis, probably through controls on pH and Eh, was optimum during periods of higher salinity.

End_of_Article - Last_Page 439------------